![](bg8.png)
test how the filter sizes effects the results of the net-
work.
8 CONCLUSION
We have demonstrated that the max-pooling layers
have high potential to aid in compressing a networks
size to operate on mobile devices. Employing max-
pooling layers allows networks to become ‘lighter’
and be trained faster, but with reduced accuracy. With
the basic network, the standard learning libraries in
networks that don’t employ the max-pooling layers
have improved results. We have implemented a deep
learning network onto a mobile platform, tested the
performance on real-world data. We have also show
that the outputs on an array of varying neural net-
works to demonstrate how the networks were ef-
fected by the max-pooling layer. By comparing the
cost/reward, we recommend the Basic max-pooling
network as it has high accuracy, with little impact on
the phones memory or processing capabilities for mo-
bile platforms.
REFERENCES
Abadi, M., Barham, P., Chen, J., Chen, Z., Davis, A., Dean,
J., Devin, M., Ghemawat, S., Irving, G., Isard, M.,
Kudlur, M., Levenberg, J., Monga, R., Moore, S.,
Murray, D. G., Steiner, B., Tucker, P., Vasudevan, V.,
Warden, P., Wicke, M., Yu, Y., Zheng, X., Brain, G.,
Osdi, I., Barham, P., Chen, J., Chen, Z., Davis, A.,
Dean, J., Devin, M., Ghemawat, S., Irving, G., Isard,
M., Kudlur, M., Levenberg, J., Monga, R., Moore,
S., Murray, D. G., Steiner, B., Tucker, P., Vasude-
van, V., Warden, P., Wicke, M., Yu, Y., and Zheng,
X. (2016). TensorFlow : A System for Large-Scale
Machine Learning. Osdi.
Alarifi, J., Goyal, M., Davison, A., Dancey, D., Khan, R.,
and Yap, M. H. (2017). Facial Skin Classification Us-
ing Convolutional Neural Networks. In Image Anal-
ysis and Recognition: 14th International Conference,
ICIAR 2017, Montreal, QC, Canada, July 5–7, 2017,
Proceedings, volume 10317, page 479. Springer.
Albert, A. M. and Ricanek Jr, K. (2008). The MORPH
database: investigating the effects of adult craniofa-
cial aging on automated face-recognition technology.
Forensic Science Communications, 10(2).
Bulat, A. and Tzimiropoulos, G. (2017a). Binarized Con-
volutional Landmark Localizers for Human Pose Esti-
mation and Face Alignment with Limited Resources.
Bulat, A. and Tzimiropoulos, G. (2017b). How far are we
from solving the 2D & 3D Face Alignment problem?
(and a dataset of 230,000 3D facial landmarks).
Cao, C., Weng, Y., Zhou, S., Tong, Y., and Zhou, K. (2014).
FaceWarehouse: A 3D facial expression database for
visual computing. IEEE Transactions on Visualization
and Computer Graphics, 20(3):413–425.
Chen, W., Wilson, J. T., Tyree, S., Weinberger, K. Q., and
Chen, Y. (2015). Compressing Neural Networks with
the Hashing Trick. Proceedings of The 32nd Inter-
national Conference on Machine Learning, 37:2285–
2294.
Chollet, F. (2016). Keras.
Co, X. M. T. (2017). BeautyPlus.
Corporation, I., Garage, W., and Itseez (2000). OpenCV.
Fagg, A., Lucey, S., and Sridharan, S. (2017). Fast , Dense
Feature SDM on an iPhone. pages 95–102.
Google (2017). Google Play Store.
Goyal, M., Reeves, N., Rajbhandari, S., Spragg, J., and Yap,
M. H. (2017). Fully Convolutional Networks for Di-
abetic Foot Ulcer Segmentation. Systems, Man, and
Cybernetics (SMC), 2017 IEEE International Confer-
ence on.
“Grother”, P. and Ngan, M. (2016). The IJB-A Face Identi-
fication Challenge Performance Report.
Hochreiter, S. and Schmidhuber, J. (1997). Long short-term
memory. Neural computation, 9(8):1735–1780.
Inc, G. (2017a). Android Studio.
Inc, S. (2017b). Snapchat.
Kendrick, C., Tan, K., Williams, T., and Yap, M. H. (2017).
An Online Tool for the Annotation of 3D Models.
pages 362–369.
Kim, Y.-D., Park, E., Yoo, S., Choi, T., Yang, L., and Shin,
D. (2016). Compression of Deep Convolutional Neu-
ral Networks for Fast and Low Power Mobile Appli-
cations. Iclr, pages 1–16.
King, D. (2013). Dlib.
Kingma, D. P. (2015). A : a m s o. pages 1–15.
Krizhevsky, A., Sutskever, I., and Hinton, G. E. (2012). Im-
ageNet Classification with Deep Convolutional Neu-
ral Networks. Advances In Neural Information Pro-
cessing Systems, pages 1–9.
Lai, H., Xiao, S., Pan, Y., Cui, Z., Feng, J., Xu, C., Yin,
J., and Yan, S. (2015). Deep Recurrent Regression for
Facial Landmark Detection. pages 1–13.
Learning, D. (2017). Deep Learning for Consumer Devices
and Services. (APRIL).
Litjens, G., Kooi, T., Bejnordi, B. E., Setio, A. A. A.,
Ciompi, F., Ghafoorian, M., van der Laak, J. A. W. M.,
van Ginneken, B., and S
´
anchez, C. I. (2017). A Survey
on Deep Learning in Medical Image Analysis. (1995).
L’Oreal (2017). Makeup genius.
Luo, P., Wang, X., and Tang, X. (2012). Hierarchical Face
Parsing via Deep Learning. pages 1–8.
Mathias, M., Benenson, R., Pedersoli, M., and Van Gool,
L. (2014). Face detection without bells and whis-
tles. Lecture Notes in Computer Science (including
subseries Lecture Notes in Artificial Intelligence and
Lecture Notes in Bioinformatics), 8692 LNCS(PART
4):720–735.
Megvii, I. (2015). Face Plus Plus.
Ranjan, R., Sankaranarayanan, S., Castillo, C. D., and Chel-
lappa, R. (2017). An All-In-One Convolutional Neural
Network for Face Analysis. pages 17–24.
VISAPP 2018 - International Conference on Computer Vision Theory and Applications
196